Mixing strand-specific and non strand-specific reads in RNA-seq

[alexanderxcy]

Hi all,

We have some old RNA-seq data sequenced from non strand-specific libraries. Recently we got some new data from the same samples sequenced from strand-specific libraries. We intended to get deeper sequencing but somehow got different preparation.

Can we mix the reads and align them as all non strand-specific reads in STAR? Would it be problematic in the Cuffdiff analysis after alignment?

Thank you!

[westerman]

While it has been a while since I did so, I did not have problems mixing strand-specific and non-strand specific libraries thus making them all non-strand-specific. You lose information about the strand, of course, but nothing else.

[NicoBxl]

And is it a good idea to perform expression analysis using both unstranded and stranded samples ? using htseq-count and DESeq for example ?